326
Chapter 30
is in preparation for the Area 3.2(2) program calling for “a system
approach to understanding the operational requirements of EO systems”.
Support is welcome in the sense of target site provision, preparation and
attendance during over-flights, resulting costs are to be paid by
corresponding local/regional authorities or research groups. This would
help to close the remaining funding gap for the HyMaS instrument and
mission development.
While the IS envisaged herein and in associated programs are rather well
covered technically, there is a variety of further mission and technical routes,
some examples are briefly addressed here, without any priority assignment:
The currently demanded spectral regimes cover always the same
regions, i.e., the VNIR, typically 400–1050 nm, and SWIR (1000–2500
nm), occasionally discrete MIR and/or TIR bands are provided.
However, technically, the spectral range can be extended towards lower
wavelengths (e.g., down to 200 nm, for atmospheric measurements), as
well as longer wavelengths, with a limit for push-broom type
instruments at about 12
for detector array reasons, e.g., for
geological/ petrological applications (NB: the CMT SWIR arrays are
derivatives from longer wavelength detector arrays!)
Though not demanded by agencies such as ESA, but by distinct users, it
would be technically rather easy to add a high-resolution panchromatic
channel or 2–3 wide band spectral channels to an IS for the a.m. optics
developments (actually: HSI had it), simply by adding a linear array with
detector pixel sizes a factor of 3–4 smaller, e.g., 10 m SSP compared to
IS hyperspectral bands 40 m
Higher spatial and/or spectral resolution is technology-limited to about
30 m at 800 km altitude (electronics); even 30 m would yield a rather
large IS which cannot be accommodated on smaller satellites. A
simplified “PRISM” is about the limit for “small-satellites”, i.e., the
requirements particularly for radiometric accuracy need to be relaxed to
arrive at compact instruments
American companies started to develop Fourier transform imaging
interferometers instead of dispersing spectrometers; we do not see the
drive for this trend unless spectral resolutions below 1 nm are demanded.
However, the European industry is preparing to also address this topic,
so far on in-house funds
Another point to be addressed by users with experience in the processing
of IS data is the general reluctance to accept “quasi loss-free” data
compression algorithms instead of perfectly loss-free. We define “quasi
loss-free” as losses in the noise of the instrument. The gain in data rate
for on-board storage and transmission would be tremendous, typically a
factor 4. We see this as essential for later operational space-borne IS,
Chapter 30
is in preparation for the Area 3.2(2) program calling for “a system
approach to understanding the operational requirements of EO systems”.
Support is welcome in the sense of target site provision, preparation and
attendance during over-flights, resulting costs are to be paid by
corresponding local/regional authorities or research groups. This would
help to close the remaining funding gap for the HyMaS instrument and
mission development.
While the IS envisaged herein and in associated programs are rather well
covered technically, there is a variety of further mission and technical routes,
some examples are briefly addressed here, without any priority assignment:
The currently demanded spectral regimes cover always the same
regions, i.e., the VNIR, typically 400–1050 nm, and SWIR (1000–2500
nm), occasionally discrete MIR and/or TIR bands are provided.
However, technically, the spectral range can be extended towards lower
wavelengths (e.g., down to 200 nm, for atmospheric measurements), as
well as longer wavelengths, with a limit for push-broom type
instruments at about 12
for detector array reasons, e.g., for
geological/ petrological applications (NB: the CMT SWIR arrays are
derivatives from longer wavelength detector arrays!)
Though not demanded by agencies such as ESA, but by distinct users, it
would be technically rather easy to add a high-resolution panchromatic
channel or 2–3 wide band spectral channels to an IS for the a.m. optics
developments (actually: HSI had it), simply by adding a linear array with
detector pixel sizes a factor of 3–4 smaller, e.g., 10 m SSP compared to
IS hyperspectral bands 40 m
Higher spatial and/or spectral resolution is technology-limited to about
30 m at 800 km altitude (electronics); even 30 m would yield a rather
large IS which cannot be accommodated on smaller satellites. A
simplified “PRISM” is about the limit for “small-satellites”, i.e., the
requirements particularly for radiometric accuracy need to be relaxed to
arrive at compact instruments
American companies started to develop Fourier transform imaging
interferometers instead of dispersing spectrometers; we do not see the
drive for this trend unless spectral resolutions below 1 nm are demanded.
However, the European industry is preparing to also address this topic,
so far on in-house funds
Another point to be addressed by users with experience in the processing
of IS data is the general reluctance to accept “quasi loss-free” data
compression algorithms instead of perfectly loss-free. We define “quasi
loss-free” as losses in the noise of the instrument. The gain in data rate
for on-board storage and transmission would be tremendous, typically a
factor 4. We see this as essential for later operational space-borne IS,